NVIDIA A40 PCIe vs NVIDIA GeForce GT 640
We compared two Desktop platform GPUs: 48GB VRAM A40 PCIe and 2GB VRAM GeForce GT 640 to see which GPU has better performance in key specifications, benchmark tests, power consumption, etc.
Main Differences
NVIDIA A40 PCIe 's Advantages
Released 8 years and 4 months late
Boost Clock1740MHz
More VRAM (48GB vs 2GB)
Larger VRAM bandwidth (695.8GB/s vs 28.51GB/s)
10368 additional rendering cores
NVIDIA GeForce GT 640 's Advantages
Lower TDP (65W vs 300W)
Score
Benchmark
FP32 (float)
A40 PCIe
+5299%
37.42 TFLOPS
GeForce GT 640
0.693 TFLOPS
Blender
A40 PCIe
+12368%
3990
GeForce GT 640
32
Graphics Card
Oct 2020
Release Date
Jun 2012
Tesla
Generation
GeForce 600
Desktop
Type
Desktop
PCIe 4.0 x16
Bus Interface
PCIe 3.0 x16
Clock Speeds
1305 MHz
Base Clock
-
1740 MHz
Boost Clock
-
1812 MHz
Memory Clock
891 MHz
Memory
48GB
Memory Size
2GB
GDDR6
Memory Type
DDR3
384bit
Memory Bus
128bit
695.8GB/s
Bandwidth
28.51GB/s
Render Config
-
-
-
84
SM Count
-
10752
Shading Units
384
336
TMUs
32
112
ROPs
16
336
Tensor Cores
-
84
RT Cores
-
128 KB (per SM)
L1 Cache
16 KB (per SMX)
6 MB
L2 Cache
256 KB
-
-
-
Theoretical Performance
194.9 GPixel/s
Pixel Rate
7.216 GPixel/s
584.6 GTexel/s
Texture Rate
28.86 GTexel/s
37.42 TFLOPS
FP16 (half)
-
37.42 TFLOPS
FP32 (float)
692.7 GFLOPS
584.6 GFLOPS
FP64 (double)
28.86 GFLOPS
Board Design
300W
TDP
65W
700 W
Suggested PSU
250 W
3x DisplayPort 1.4a
Outputs
1x DVI
1x HDMI 1.4a
1x DisplayPort 1.2
8-pin EPS
Power Connectors
None
Graphics Processor
GA102
GPU Name
GK107
-
GPU Variant
GK107-300-A2
Ampere
Architecture
Kepler
Samsung
Foundry
TSMC
8 nm
Process Size
28 nm
28.3 billion
Transistors
1.27 billion
628 mm²
Die Size
118 mm²
Graphics Features
12 Ultimate (12_2)
DirectX
12 (11_0)
4.6
OpenGL
4.6
3.0
OpenCL
3.0
1.3
Vulkan
1.1
8.6
CUDA
3.0
6.6
Shader Model
5.1
